Overclock.net banner

4581 - 4600 of 13064 Posts

·
Registered
Joined
·
463 Posts
Hmmm, I can't really remember. I use an offset with my method, but it was around +.137V. I do remember that I had to put the mobo in LN2 mode to get it to boot/behave with the voltage offset I needed. It's the first time I attempted to use that much voltage and haven't had cold enough weather to try it again. I was wearing shorts and a tee today here in Texas.

Edit:
I wanna say it ended up around a high 1.5V to low 1.6V.

Edit 2:
Unfortunately, that screen doesn't show during the actual run. I didn't open HWiNFO and CPUZ until after the run was completed. So the voltage it shows wasn't what it actually ran at.
How the new agesa is going with memory overclock?
 

·
Registered
Joined
·
181 Posts
How the new agesa is going with memory overclock?
It is a better revision imho. I have lowered the dram back to 1.35v, SOC is down by .05v to 9.43v(havent tested this any further yet), CPU is down to offset+0.25v from 0.5v with no loss obvious. I have managed to boot and run testing at 3600 which failed hcimemtest @ 50%. This wasn't a very
hardy attempt. I am short on time for next few weeks. I am not seeing any glitches so far.

I would say go for it, if you are unsure.
 

·
Computer Enthusiast
Joined
·
2,919 Posts
Ditto

I think this build is just the corrected 1101 - seems good so far. AGESA 1006 is an improvement so far, but more thorough testing ahead :)
 

·
Registered
Joined
·
202 Posts
1103 let's me run at 3533 on my 3466 timings now even if my temps are 43c on ram. Couldn't do this at all on previous. Only other thing I've changed was 1.425v to 1.44v. I left my PC running all night and no error. So happy 🙂 think it's the max I can go on this CPU imc though
 

·
Meddling user
Joined
·
7,423 Posts
Possibly a mixture of the two in order to condense the Precision stuff. I'll see if I can find out.
Cheers appreciate it :) .

What is still so shameful is SOC frequency in manual mode still has bugged out range :( ; this has been going on for months.

Ditto

I think this build is just the corrected 1101 - seems good so far. AGESA 1006 is an improvement so far, but more thorough testing ahead :)
I agree this seems like sorted 1101.

For me jury is out if I prefer AGESA 1.0.0.6 UEFIs or 1.0.0.2C, I've gone back to 32GB and really can't say it's any better for supporting it. As the older AGESA allows manual tweaks of PPT, TDC, EDC and Scalar value I'm leaning towards going back TBH.
 

·
Registered
Joined
·
49 Posts
Hey all, I've been lurking this thread for a few days since I heard the new AGESA was coming out, I'd like to cc ya'll to this post I made in the DDR4 thread https://www.overclock.net/forum/10-...memory-stability-thread-206.html#post27739694
I'm wondering about C7H-specific stable results for you guys as far as bdie goes, and where I could improve.

Also it seems like 1.0.0.6 has the same or worse dual rank compatibility compared to 1.0.0.2? What a time for me to pick up a 2x16 kit lol.
 

·
Premium Member
Joined
·
2,754 Posts
1103 bios or AGESA 1.0.0.6 in general seems to include new PMU FW versions for all Ryzen CPUs (Summit, Raven / Pinnacle) so this time around any changes in the memory behavior might actually be based on facts, instead of placebo.

It is great if there are gains to be had from improving the memory controller firmwares, however at the same time it would be quite sad: If the behavior can be improved, why has it taken so long to do so? AFAIK the previous firmware versions date from Raven's release, i.e. are already 10 months old...
 

·
Registered
Joined
·
2,316 Posts
Has anyone else made further attempts to get 3600 stable at 14-14-14-14? I can get further in testing before errors, but can't complete testing error free yet.

And I still haven't pinned down what caused my Aida latency to increase 2ns from when I first flashed 1103. I know it's minuscule, but 57ns latencies was something I could repeatedly achieve and now I can't get lower than 59ns again.
 

·
Computer Enthusiast
Joined
·
2,919 Posts
1103 bios or AGESA 1.0.0.6 in general seems to include new PMU FW versions for all Ryzen CPUs (Summit, Raven / Pinnacle) so this time around any changes in the memory behavior might actually be based on facts, instead of placebo.

It is great if there are gains to be had from improving the memory controller firmwares, however at the same time it would be quite sad: If the behavior can be improved, why has it taken so long to do so? AFAIK the previous firmware versions date from Raven's release, i.e. are already 10 months old...
Interesting bit of information - I had not noticed that! My testing is very very basic so far - i'm going to spend a few hours doing my OC from (almost) scratch to see if it's real improvement though. Currently i'm running 3533 CL14 13 13 13 with tight timings, but SoC voltage is slightly lower than previously needed, but I've not tested enough yet. I did a little experimenting at 3600 (which I did have stable on 0804) but previous settings do not work.

Someone else (maybe more) reported their OC on 1101 had just stopped working without any obvious reason. Same for me, I had tested my 3600 fairly extensively and had been running it for a couple of weeks when it just.... well, it just broke - I cannot explain it any other way. I considered that maybe it was a problem with Windows, but it's doesn't appear to be the culprit. It's really stumped me for now.

I'm going to start with OC of 3200, with geardown=off and go from there....

Yes, frustrating if they could have done this all along - I do have a nagging feeling they might not have put as much work into the PMU firmware as they should - we'll see. Still, any improvements are welcome.
 

·
Computer Enthusiast
Joined
·
2,919 Posts
Hmmmm.... first 3533 OC attempt with bios 1103

I put in my previous day to day settings for 4.1GHz 3533MTs CL14 - same settings I was booting and running before the update - but they do NOT work. The CPU side of things works fine - lovely stable 4.1, but ram - doesn't want to know, I haven't booted once with my last ram settings - even with no CPU OC.

I will report back if and when I learn more.....
 

·
Overclocker
Joined
·
7,030 Posts
Is it save to flash this "stable" BIOS without it crapping out drivers like the previous one did?
 

·
Computer Enthusiast
Joined
·
2,919 Posts
Is it save to flash this "stable" BIOS without it crapping out drivers like the previous one did?
I can only speak for myself, but yes, that problem is resolved. ROG Supreme FX has returned.

Not had enough time to properly check out the memory OC side of things, but the settings I had working yesterday didn't boot at all. I did NOT test extensively and will star from scratch tomorrow. I may re-check those settings before I crash out tonight, just to stop me from wondering about that :)

Edit: I really do think it would be MUCH better for the beta bios to be tested out in the enthusiast community long enough for us to give reliable, tested feedback.
 

·
Computer Enthusiast
Joined
·
2,919 Posts
So, I gave my previously working settings from bios 1101 - 4.1GHz 3533 - another shot with bios 1103... I loaded defaults and completely booted before trying this.

Again, it failed to boot (mem training, I I set it to do up to 3 retries) at all - tried changing a couple of things but no difference. I don't know if this is a problem, or just some setting that is working differently on AGESA 1006, but it's something that (at least for me) requires a bit more investigation. Maybe it's particular to 1xxx CPUs. Might try re-flashing, but I did flash it using afuefix.efi which does clear all settings that may linger from previous bios.

As previously mentioned, 4.1GHz CPU only OC is fine.

Very much looking fwd to seeing what results others get with this bios so far.
 

·
Banned
Joined
·
403 Posts
Crosshair VII

Crosshair VII is here, and plenty of feedback from OCN has made it into the product. It's not revolutionary, but there are a lot of small little tweaks that in the end makes a big difference.


Key highlights:

- CPU/SOC VRM reconfigured to 10+2 and upgraded to IR3555 Power Stages. Additionally the SOC phases are placed between the CPU phases for lower temperatures.

- Supports Asynchronous eCLK mode, which allows you to overclock the CPU reference clock independently from Fabric/DRAM and PCI-E.

- Performance Enhancer option can increase or disable XFR power and current limits allowing you to boost higher and longer. Together with reference clock adjustments this means you can get up to ~4.5 GHz in single threaded loads and 4.2-4.3 GHz in multi threaded. @The Stilt is the original author of this function.

- Much improved voltage monitoring both in software and at ProbeIt points with differential sensing for CPU Core Voltage, CPU SOC Voltage, DRAM Voltage and EPS12V.

- Stealth mode which easily disables Aura and on-board LEDs completely, including the power button and Q-Code display.


For further details, check the attached PDF brief. Or ask in the thread :p


BIOS

Latest releases available here: http://www.mediafire.com/folder/jkoxu4r003qu7/Release

C7H/WIFI BIOS 0601

- First release on OCN

http://www.mediafire.com/file/lyc7if2402j4203/ROG-CROSSHAIR-VII-HERO-ASUS-0601.zip sha256 d0d54fe84bf4cacc6cce5abbb071f83166ab4a4ada3d97749632cde1aa50b1e6
http://www.mediafire.com/file/dhf3sqmbg4sh3pj/ROG-CROSSHAIR-VII-HERO-WIFI-ASUS-0601.zip sha256 28bbc18cf2ec9166feb3776eaffee1d4c17f02e59d11f542bb6568071e1011fb

See you went all out on the Mobo's again your build. I thought about it but just went with a pretty entry level MSI x470, hoping to at least hit 4.2ghz with the 2600x and should I think anyways...just nothing Fancy. Good Reliable backup and Redundant build.
 

·
Overclocker & Enthusiast
Joined
·
143 Posts
So, I gave my previously working settings from bios 1101 - 4.1GHz 3533 - another shot with bios 1103... I loaded defaults and completely booted before trying this.

Again, it failed to boot (mem training, I I set it to do up to 3 retries) at all - tried changing a couple of things but no difference. I don't know if this is a problem, or just some setting that is working differently on AGESA 1006, but it's something that (at least for me) requires a bit more investigation. Maybe it's particular to 1xxx CPUs. Might try re-flashing, but I did flash it using afuefix.efi which does clear all settings that may linger from previous bios.

As previously mentioned, 4.1GHz CPU only OC is fine.

Very much looking fwd to seeing what results others get with this bios so far.
Nope I have exactly the same issue memory side, I was very close to getting 3533MHz stable on my 2700X on bios 1002 infact, going to 1103 before I even had a thorough chance to test 1101 and I can't even get close no matter what I change. It won't even post 3533MHz at this stage, where as previously I could at least boot into 3666MHz (not stable ofc). I have tried ProcODT at 40 all the way to 80 each yield same results as well as the other various terminations to 20, 24 and then 20, 20, 40, 60 and no dice. SOC voltage at 0.93 actually gave me the closest to stability I actually got into windows and passed HCI up until 50%, while the rest didn't even post all together. I had memory retry to 7 even, so I was a bit more patient for nothing basically.
 

·
Banned
Joined
·
403 Posts
Nope I have exactly the same issue memory side, I was very close to getting 3533MHz stable on my 2700X on bios 1002 infact, going to 1103 before I even had a thorough chance to test 1101 and I can't even get close no matter what I change. It won't even post 3533MHz at this stage, where as previously I could at least boot into 3666MHz (not stable ofc). I have tried ProcODT at 40 all the way to 80 each yield same results as well as the other various terminations to 20, 24 and then 20, 20, 40, 60 and no dice. SOC voltage at 0.93 actually gave me the closest to stability I actually got into windows and passed HCI up until 50%, while the rest didn't even post all together. I had memory retry to 7 even, so I was a bit more patient for nothing basically.
Well you have the Tuning and Power delivery with ROG there. Surprised not at 4.2ghz. And memory Experimentation, hell I don't know some kits can oc really well. Me I don't care I thought 3200mhz was still Nice. Shot in the dark to see if XMP is gonna work or not, it probably does though. And I salute them..I thought Ryzen was pretty Worthy. I'll have a mildly operated one and see what I can play with and find lol
 

·
Overclocker & Enthusiast
Joined
·
143 Posts
Well you have the Tuning and Power delivery with ROG there. Surprised not at 4.2ghz. And memory Experimentation, hell I don't know some kits can oc really well. Me I don't care I thought 3200mhz was still Nice. Shot in the dark to see if XMP is gonna work or not, it probably does though. And I salute them..I thought Ryzen was pretty Worthy. I'll have a mildly operated one and see what I can play with and find lol
My processor speed is 4.265GHz all core, single core is 4.5GHz. You probably got mixed up with the guy with the first gen chip who has 4.1GHz (which is very good). My kit is samsung b-die, single rank 2x8GB which is pretty much the best you can get for the IMC so being locked at 3466 CL14 is kinda sad, when the kit it's self can do 3700 CL14 on a CPU with a capable IMC. Zen 2 looks to be promising for that side though so, we'll see. I'll aim low and hope I can actually do 3700MHz CL14, high end I'll try for around CL15/Cl16 4000MHz 1T but it's all gonna depend on the new IMC on those chips, anyway I'm going off on a tangent now haha. FYI if you didn't see the signature the kit I'm using right now is rated for 4600MHz CL19 (2T) so hence why it's a little sad to see such low memory speeds in comparison to what it could be doing.
 

·
Overclocker
Joined
·
7,030 Posts
I can only speak for myself, but yes, that problem is resolved. ROG Supreme FX has returned.

Not had enough time to properly check out the memory OC side of things, but the settings I had working yesterday didn't boot at all. I did NOT test extensively and will star from scratch tomorrow. I may re-check those settings before I crash out tonight, just to stop me from wondering about that :)

Edit: I really do think it would be MUCH better for the beta bios to be tested out in the enthusiast community long enough for us to give reliable, tested feedback.
Thnx dude, the only issue i had was with the drivers other than that the agesa 6 is pretty good as far as i could tell.

I thought this was going to be an quiet Sunday but it turns out that i have a lot of things to do now with this new BIOS lol.
 

·
Banned
Joined
·
403 Posts
My processor speed is 4.265GHz all core, single core is 4.5GHz. You probably got mixed up with the guy with the first gen chip who has 4.1GHz (which is very good). My kit is samsung b-die, single rank 2x8GB which is pretty much the best you can get for the IMC so being locked at 3466 CL14 is kinda sad, when the kit it's self can do 3700 CL14 on a CPU with a capable IMC. Zen 2 looks to be promising for that side though so, we'll see. I'll aim low and hope I can actually do 3700MHz CL14, high end I'll try for around CL15/Cl16 4000MHz 1T but it's all gonna depend on the new IMC on those chips, anyway I'm going off on a tangent now haha. FYI if you didn't see the signature the kit I'm using right now is rated for 4600MHz CL19 (2T) so hence why it's a little sad to see such low memory speeds in comparison to what it could be doing.
Yeah seems like they wanted B-Die on everything and probably much more of the testing, But that Microcenter stuff was a decent Hynix kit on sale, Think I did get it up to 3400-3600mhz. Just ok, nothing to write home about. Someone said it was Working with Ryzen so good to know. cool stuff. Save Team Blue for a Rainy day been wanting to play with Ryzen for some time..2nd Gen seems quite Nice. Only thing I didn't order the 2600x with the Spire cooler...kinda wish I did now I think about it but o well, going under water.
 

·
Computer Enthusiast
Joined
·
2,919 Posts
Nope I have exactly the same issue memory side, I was very close to getting 3533MHz stable on my 2700X on bios 1002 infact, going to 1103 before I even had a thorough chance to test 1101 and I can't even get close no matter what I change. It won't even post 3533MHz at this stage, where as previously I could at least boot into 3666MHz (not stable ofc). I have tried ProcODT at 40 all the way to 80 each yield same results as well as the other various terminations to 20, 24 and then 20, 20, 40, 60 and no dice. SOC voltage at 0.93 actually gave me the closest to stability I actually got into windows and passed HCI up until 50%, while the rest didn't even post all together. I had memory retry to 7 even, so I was a bit more patient for nothing basically.
Interesting - we have similar kits - I'm just about to star from scratch see if I can find whats changed, or if 3533 upwards are not working, something @Silent_Scone should be aware that we're checking out.

Again - why didn't ASUS leave this with enthusiasts for more than a couple of days before general release? I'm going to start by re-flashing, but I'm not holding my breath, Stilt mentioned the new firmwares including the IMC so you're my hunch (like others) is that we need to re-tune. Otherwise some of us are having problems between v1101`and 1103. Maybe this re-flash will help....here I go! :rambo:
 
4581 - 4600 of 13064 Posts
Top